- Email: [email protected]
Laser Droplet Generation from a Metal Foil
Available online at www.sciencedirect.com
ScienceDirect Physics Procedia 56 (2014) 720 – 729
8th International Conference on Photonic Technologies LANE 2014
Laser droplet generation from a metal foil Andrej Jeromena,*, Alexander Kuznetsova, Edvard Govekara a
Faculty of Mechanical Engineering, University of Ljubljana, Aškerþeva 6, SI-1000 Ljubljana, Slovenia
Abstract Metal droplets are used in various droplet based innovative technologies, especially in joining of temperature sensitive components where the amount of applied energy and its spatial distribution are important. Laser light has already been proved to be a suitable energy source for metal droplet generation, since it enables high spatial and temporal control of the energy input. In this contribution, a novel process of laser droplet generation (LDG) from a metal foil is presented, which could provide increased flexibility of the generated droplet parameters compared to the existing LDG processes. In the novel LDG process, a horizontally placed foil is irradiated by a laser beam from above and the melted part of the foil is detached by shielding gas overpressure above the foil. The results of presented preliminary experimental study show that droplet formation from metal foil is possible by means of an annular laser beam. In addition to this, the space of process parameters is explored, showing influences of temporal intensity distributions of the annular laser beam, shielding gas overpressure, and foil thickness and material on the process outcome. © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and blind-review under responsibility of the Bayerisches Laserzentrum GmbH. Peer-review under responsibility of the Bayerisches Laserzentrum GmbH Keywords: laser droplet generation; metal foil; annular laser beam
1. Introduction Nowadays, metal droplets are used as a basis of various innovative technologies, e.g. Chun and Passow (1993) and Liu (2000), which are employed in different applications. Some of the reported applications, based on metal droplets, include joining of electrical contacts by Dreizin (1997) and Conway et al. (2002), joining of heat sensitive and dissimilar materials by Jahrsdörfer et al. (2003) and Govekar et al. (2009), welding of zinc coated steel sheets by
* Corresponding author. Tel.: +386-1-4771-514; fax: +386-1-251-8567 . E-mail address: [email protected]
1875-3892 © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Bayerisches Laserzentrum GmbH doi:10.1016/j.phpro.2014.08.079
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
721
Jeriþ et al. (2009), microcasting by Zarzalejo et al. (1999), 3-D structuring by Qi et al. (2012), and droplet brazing by Held et al. (2012) and Jeromen et al. (2014). For different applications, different droplet generation systems and processes have been developed. They could be divided into two groups: drop-on-demand, where individual droplets are separately generated, and continuous droplet generation, where a sequence of identical droplets is generated at a constant frequency. Drop-on-demand process is required in applications, where metal droplets need to be deposited individually to realize point-like joints, for example in electrical contacting of temperature-sensitive electrical and electronic components such as piezoelectric laminates, described by Albert et al. (2011). Originally, metal drop-on-demand generation process with solid metal as input material was introduced by Dreizin (1997) who generated droplets by melting a metal wire using an electric arc. Later, laser droplet generation from a wire (LDGw) has been described by Hoving and Jahrsdörfer (2001) and analyzed by Kokalj et al. (2006). In LDGw, the droplet is generated by melting the tip of a metal wire by a laser pulse. As laser light enables high spatial and temporal control of the energy input it proved to be a very suitable energy source for metal droplet generation. However, to reliably generate metal droplets from a wire, a symmetrical energy input by means of an annular laser beam, as well as a precise control and synchronization of laser pulse power and wire feeding are needed, as demonstrated by Kuznetsov et al. (2014). Recently, Held et al. (2012) reported on laser droplet brazing, where a spherical braze preform, blocking a tapered nozzle, is melted by the laser pulse and detached from the nozzle by the overpressure of shielding gas inside the nozzle. The technique is promising for electrical contacting applications, if the nozzle instability issue due to the thermal shocks is addressed adequately. As an alternative laser droplet generation process, a novel process of laser droplet generation from a foil (LDGf) is presented in this paper. The LDGf process has a potential to provide increased flexibility of the generated droplet parameters, in particular generation of smaller droplets, compared to the above laser droplet generation processes by using a less complex system. What is more, in LDGf, in contrast to laser droplet brazing system, no permanent system components are thermally loaded which implies higher stability of the system. In the following, the LDGf experimental system and process are described. Next, the results of the initial experimental study of LDGf, including two different laser beam shapes, two different foil materials and different foil thicknesses, are presented and discussed. In the conclusions, the main results of the study are gathered. 2. Experimental system and process of laser droplet generation from a foil The LDGf experimental system is presented in Fig. 1(a). It consisted of a Nd:Yag pulsed laser source with wavelength of 1064 nm and a Gaussian beam. Optionally, an optical system, designed by Jeriþ (2010), was used to shape the Gaussian laser beam into an annular laser beam (shown in Fig. 1(a)). The approximate caustic of the annular laser beam, determined by measuring the inner and outer diameter of the heat affected zone on irradiated metal surface at different positions along the beam axis, is presented in Fig. 1(b). Circles in Fig. 1(b) denote the outer and squares the inner diameter of the annular laser beam. The laser beam was guided to the surface of the metal foil. For this purpose, an experimental cell, pressurized by shielding gas, was designed as shown in Fig. 1(a). The top and the bottom opening of the cell were closed by a glass window, transparent for the laser light, and by the clamped metal foil. The overpressure p of argon shielding gas inside the cell was set manually and was monitored by a pressure transducer. The volume of the cell was approximately 4 cm3. To prevent oxidation, the argon atmosphere was maintained also below the foil. For the experimental process characterization, a high speed IR camera was used. The LDGf process consists phenomenologically of two consecutive phases: (1) laser beam selective melting of the foil, and (2) detachment of the melted material by the shielding gas overpressure with formation of the droplet due to surface tension. At given metal foil thickness h and its physical properties, the main process parameters are the spatial profile (shape) of the laser beam, irradiated spot diameter ds, laser pulse power time profile P(t) and shielding gas overpressure p. The foil material used in experiments was nickel of purity 99.0% and thickness h = 0.1 and 0.0125 mm. In view of potential brazing applications in electrical contacting of copper wires on silver contact pads, the Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was also selected as it is used for flux-free brazing of copper. The physical properties of both materials are gathered in Table 1. Instead of the unavailable data for Ag15Cu80P5, the values for copper and
722
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
silver are given. Compared to Ag15Cu80P5, Ni has a higher melting point and, considering values for copper and silver, much lower heat conduction, higher specific heat and higher viscosity.
Fig. 1. (a) Scheme of LDGf experimental system; (b) Annular laser beam caustic.
Table 1. Physical properties of used foil materials (from Liu (2000) except where indicated). Property
Ni
Ag15Cu80P5
Density ȡ at 300 K [kg/m3]
7905
8442 (product data sheet)
Temperature of melting Tm [K]
1728
917 (Ts), 1073 (Tl) (product data sheet)
Thermal conductivity O at 300 K [W/(m·K)]
88.5
397 (Cu), 425 (Ag)
Specific heat cp at 300 K [J/(kg·K)]
452
386 (Cu), 234 (Ag)
Viscosity ȝ at melting point [Pa·s]
1.7·10-4
3.0·10-4 (Cu), 4.5·10-4 (Ag)
Surface tension V at melting point [N/m]
1.8
1.3 (Cu), 0.90 (Ag)
3. Experiments, results and discussion In the following, the results of the initial LDGf experiments and the influence of the main process parameters on the process outcome are presented and discussed. In experiments, the Gaussian and the annular laser beam were used at different diameters ds of irradiated spot. In the experiments also different values of the shielding gas overpressure p were applied, and nickel and Ag15Cu80P5 alloy foils of selected thicknesses were used. The process was characterized by the high speed IR camera with frame of 96x80 pixels, integration time of 6 ȝs and frame rate of 1660 fps. 3.1. Gaussian laser beam First, the LDGf experiments were performed using the Gaussian laser beam, which was selected as the most common laser beam profile. To select different values of irradiated spot diameter ds, the experimental cell with the foil was positioned at the corresponding position above the laser beam focus. By the foil being positioned above the
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
723
laser beam focus, additional heating of the detached foil material, which travelled downwards through the laser beam focus, was realized. In the experiments, Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was used with two irradiated spot diameter ds values ds = 1.0 and 2.0 mm. At ds = 1.0 mm, a square laser pulse of power P = 3.0 kW and duration W = 6.5 ms and at ds = 2.0 mm a square pulse of P = 3.9 kW and W = 20.0 ms were used. The threshold shielding gas overpressure ptr for detachment of the melted part of the foil at ds = 1.0 and 2.0 mm was experimentally determined to be ptr = 8.5 kPa and 5.0 kPa, respectively. It was observed from the recorded sequences of the IR intensity images that at overpressures p above the threshold overpressure ptr no successful LDGf, resulting in a droplet, was realized. In all observed cases, the melted region either broke up and the melted material was pulled to the edge or it inflated into a bubble which subsequently disintegrated into splashes, allowing the pressurized shielding gas to escape from the cell. An example sequence of recorded IR intensity images at ds = 2.0 mm, showing unsuccessful LDGf with bubble inflation and disintegration, is presented in Fig. 2. Indicated time t below images is related to the time of the laser pulse trigger signal. In the images, brighter color denotes higher temperatures. In the leftmost image of Fig. 2, the irradiated spot is visible. After the irradiated part of the foil material was melted, the melted region was inflated into a bubble at t = 19.3 – 19.9 ms which then detached at t | 20.5 ms. After detachment, the bubble disintegrated into splashes although the laser beam was turned off at t = 20.0 ms. At the top of the rightmost two images, the circular opening that remained in the foil after detachment of the irradiated foil spot is clearly visible.
Fig. 2. Sequence of IR intensity images of an unsuccessful LDGf.
In the performed experiments with the Gaussian laser beam no successful droplet generation was observed. The process resulted into break-up of the melted region with the melt being pulled to the edges of the opening or into the bubble formation and disintegration as well as. Both outcomes were attributed to the Gaussian laser beam intensity distribution, where the highest intensity is in the center of the irradiated spot. This caused the foil in the centre of the irradiated spot to melt first and thus gave rise to the above described behavior which results in unsuccessful process outcome. 3.2. Annular laser beam To avoid melting of the irradiated region in its centre, the annular laser beam, with a ring like shape of the irradiated foil surface region, was employed. As in the Gaussian laser beam case, the experimental cell was positioned above the laser beam focus, to select the value of the irradiated spot diameter ds. This layout also enabled heating of the detached material below foil by the laser beam. The annular laser beam caustic (see Fig. 1(b)) together with the geometry of the experimental cell allowed the available values of beam outer diameter dout, which is equal to the irradiated spot diameter ds, between 1.5 and 3.0 mm. Below dout = 1.5 mm, the beam shape was not annular but full (see Fig. 1(b)), and above dout = 3.0 mm, the beam was too wide to fit the experimental cell opening. In experiments, both Ni and Ag15Cu80P5 foils were used, the former with thicknesses h = 0.1 mm and 0.0125 mm and the latter with thickness h = 0.2 mm. 3.2.1. Nickel foil with thickness of 0.1 mm For LDGf with Ni foil with thickness h = 0.1 mm, three values of irradiated spot diameter ds were selected: ds = 1.5, 2.0, and 3.0 mm. Although the LDGf using the square laser pulse profile was already successful, the results shown in the following were obtained using the laser pulse profile P(t) with decreasing power. With such profile,
724
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
faster melting of the foil without evaporation was achieved. At ds = 1.5 mm, the pulse power P linearly decreased from 8 kW to zero in duration of the pulse W = 3 ms. At ds = 2.0 mm, the pulse power P linearly decreased from 8 kW to zero in W = 5 ms. At ds = 3.0 mm, the area of the irradiated surface was so large and the corresponding laser light intensity so low, that in order to provide the necessary energy for melting, the laser pulse consisted of a constant part with P = 7.5 kW and W = 5 ms followed by a linear decrease from P = 7.5 kW to zero in the following millisecond. At the above described parameters, the second phase of LDGf process – detachment of the melted part and formation of the droplet – was observed to consist of three consecutive steps. The observed steps are presented by the sequence of IR intensity images in Fig. 3. Indicated time t below the images is related to the time of the laser pulse trigger signal. The second phase of LDGf process started after the ring shaped irradiated region of the foil was melted at t | 3.5 ms (see Fig. 3). After this, at time t = 4.7 ms, the central circular part, which was still solid, was pressed downwards due to the deformation of the molten ring by the shielding gas overpressure p. Then, the solid central part was detached by the break-up of the molten ring (t | 5.3 ms) which had connected the central part with the foil. Finally, the still solid central part, travelling downwards through the laser beam focus, was melted by the laser light and under influence of the surface tension, a droplet was formed (t = 10.7 ms).
Fig. 3. Sequence of IR intensity images of a successful LDGf at ds = 3.0 mm and p = 1.7 kPa.
The above described detachment and formation of the droplet was observed to be very repeatable and accompanied with minor or no splashes. However, depending on the shielding gas overpressure p and the irradiated spot diameter ds, different regimes of the LDGf process, leading to different outcomes, were observed experimentally. The regimes are presented in a form of an overpressure p versus diameter ds diagram, shown in Fig. 4. In the diagram, stars denote the values of p and ds where the droplet generation was observed to be optimal, i.e. repeatable and with minimum amount of splashes. At increased overpressure p, the amount of splashes, produced by the ring break-up, increased, which is denoted by squares in the diagram. At slightly lower than optimal overpressure p values, denoted by circles, the detachment was observed to become unstable. The inspection of the acquired IR image sequences showed that the instability was caused by the incomplete and un-simultaneous breakup of the molten ring which resulted in the detached material falling at an angle and, therefore, missing the laser beam focus. The described instability was attributed to the imperfect spatial symmetry of the annular laser beam which caused uneven heating and melting of the irradiated ring-shaped region of the foil. At further decreased values of the shielding gas overpressure p, denoted by crosses, the detachment did not occur at all. In those cases the melted region of the foil either remained partly or completely attached to the solid edge of the foil and resolidified.
725
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
Fig. 4. Shielding gas overpressure p versus irradiated spot diameter ds. The symbols denote different detachment regimes. Theoretical gas overpressure threshold ptr dependence is denoted by solid line.
The threshold shielding gas overpressure ptr for reliable detachment of the melted part of the foil at ds = 1.5, 2.0, and 3.0 mm was experimentally determined to be ptr = 5.5, 3.5, and 1.7 kPa, respectively. These values are denoted by star symbols in Fig. 4. By considering a simple force balance model, the threshold overpressure ptr dependence on the irradiated spot diameter ds was determined also theoretically. The model describes the force balance of the central circular part at the moment directly before detachment, when the central circular part is connected with the foil by the liquid in a shape of a circular wall (see Fig. 3 at t = 4.7 ms). In the model, the force of gravity Fg and the overpressure force Fp, acting downwards on the central circular part, are balanced with the surface tension force FV which pulls the central part upwards. The force of gravity was expressed by Fg = Sds2hȡg, where h is the foil thickness, ȡ the density of the foil material, and g the gravitational acceleration. The overpressure force was expressed by Fp = Sds2ptr/4, and the surface tension force by FV = 2SdsV, where V is the surface tension of liquid material and where both (inner and outer) surfaces of the circular wall were taken into account. The threshold overpressure ptr dependence on the diameter ds is then determined from the force balance equation Fg + Fp = FV: ptr (d s )
8V hU g . ds
(1)
It can be shown that the second term on the right, resulting from the force of gravity, is negligible since it is three orders of magnitude smaller than the first term at parameter values used in the experiments. This shows that the gravity has a negligible effect on the LDGf process. The ptr(ds) dependence, expressed by Eq. (1), is presented in Fig. 4 as a solid line and corresponds well to the experimentally determined threshold shielding gas overpressure ptr values, denoted by stars. The generated droplet diameter in general depends on the diameter ds of irradiated spot. At given diameter ds, the diameter dex of the experimentally generated droplets was determined by measuring generated droplets which solidified in a free fall. The measurement error due to the imperfect spherical shape of the solidified droplets was estimated to be 0.1 mm. The measured dex values at the selected values of the irradiated spot diameter ds are gathered in Table 2. There they are compared to the theoretically determined droplet diameter dth values, calculated from the volume of the melted spot with diameter ds of the foil by equation: d th
3
3d s 2 h . 16
(2)
It can be seen from Table 2 that the experimental values are equal to the theoretical values which indicates that, within the measurement error, the complete volume of the melted material was detached in the form of a droplet.
726
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729 Table 2. Measured dex and calculated dth generated droplet diameters at selected irradiated spot diameter ds values. Irradiated spot diameter ds [mm]
measured dex [mm]
theoretical dth [mm]
1.5
0.7 ± 0.1
0.70
2.0
0.85 ± 0.1
0.84
3.0
1.1 ± 0.1
1.11
In a view of potential joining applications, deposition of generated droplets was demonstrated on a titanium substrate of 0.6 mm thickness which was placed 6 mm below the bottom surface of the foil. In Fig. 5, photographs of the top and the side view of the deposited droplets at the selected irradiated spot diameter ds values are shown. Despite the relatively large distance between the foil and the substrate, the generated droplets contained enough energy to wet well the substrate surface.
Fig. 5. Top and side view photographs of deposited droplets at the selected values of the irradiated spot diameter ds.
3.2.2. Nickel foil with thickness of 0.0125 mm To be able to produce droplets of smaller diameter, the LDGf process was performed using thinner Ni foil with thickness h = 0.0125 mm at the irradiated spot diameter ds values of 1.5, 2.0, and 3.0 mm. At this foil thickness the experiments of the LDGf were unsuccessful at all values of the shielding gas overpressure p, regardless of the irradiated spot diameter ds and the laser pulse power profile P(t). At lower values of the shielding gas overpressure p, including p = 0, the melted part of the foil was immediately pulled by the surface tension to the edge of the opening. At increased overpressure p values, the LDGf process resulted in unstable detachment. With further increase of the overpressure p, the droplet detachment phase was accompanied with splashes and with too high acceleration of the central part to be melted while travelling through the laser beam focus. Example IR intensity sequences of the corresponding LDGf processes with no detachment and with unstable detachment are shown in Fig. 6. Indicated time t below images is related to the time of the laser pulse trigger signal. In both cases, the irradiated spot diameter ds was 2.0 mm, and the laser pulse power P linearly decreased from 2 kW to zero in W = 1.2 ms. In the first case, shown in Fig. 6(a), the overpressure p was set to 2.0 kPa. The IR images show the melted material being pulled to the edge, accompanied by some splashes. In attempt to achieve a stable detachment, the overpressure p was increased to 6.0 kPa, which is nearly twice the optimal value as determined by the theoretical model. As shown by the acquired IR images in Fig. 6(b), the detaching central part was held at one side (see image at t = 2.3 ms), detached at a large angle with respect to the vertical axis, and completely missed the laser beam focus. The main reason for the observed behavior was presumably the annular laser beam asymmetry. The beam asymmetry is evident in the first image (at t = 1.1 ms) in both Fig. 6(a) and (b), where beside the ring shaped heated region of the foil also a part of the inner circular region was heated. Due to the small thickness of the foil, the shape of the heated region approximately corresponds to the shape of the laser beam. The observed asymmetry of the laser beam was a consequence of imperfect alignment of the beam shaping optical components. Due to the resulting asymmetric melting, the central part was pulled to a side by the surface tension. The small thickness of the foil contributed considerably to the described behavior, as the melted region was evidently unstable and broke immediately. Its width to thickness ratio was approximately 50 in the case of foil thickness h = 0.0125 mm, compared to approximately 5 in the case of h = 0.1 mm, where the ring shaped liquid region was more stable and
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
broke-up only when it was thinned enough by stretching due to the central part being pressed downwards by the overpressure p.
Fig. 6. Sequence of IR intensity images of an unstable detachment at (a) p = 2.0 kPa and (b) p = 6.0 kPa.
3.2.3. Ag15Cu80P5 alloy with thickness of 0.2 mm In view of potential brazing applications the Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was employed in LDGf. As above, the irradiated spot diameter ds values of 1.5, 2.0, and 3.0 mm were selected while the corresponding laser power profiles P(t) and shielding gas overpressure p values were attempted to be experimentally determined in order to achieve a successful LDGf process. The acquired IR image sequences showed that at irradiated spot diameter ds values of 1.5 and 2.0 mm, the process outcome was very similar to the case with Gaussian beam. Despite the ring shaped irradiated region, the central region melted virtually at the same time as the irradiated region, which caused inflation of the melted region into a bubble and its disintegration into splashes, as described in Section 3.1. In attempt to avoid premature melting of the central region before the beginning of the detachment phase, the pulse power was increased to decrease time needed for melting the ring shaped irradiated region. However, before reaching fast enough melting of the irradiated region by the increased laser pulse power, the material at the irradiated surface started to evaporate, causing violent splashes. Only in the case of the irradiated spot diameter ds = 3.0 mm it was possible to melt the ring shaped region of the foil before the central region melted. Such a case, with the laser pulse power P linearly decreasing from 8 kW to zero in W = 8.0 ms and the shielding gas overpressure p = 2.0 kPa, is presented in Fig. 7(a). Indicated time t below images is related to the time of the laser pulse trigger signal. The images in Fig. 7(a) show the downward acceleration of the central part. However, the liquid, connecting the central part to the foil, did not break-up as fast as in the corresponding case with Ni (see Fig. 3). Indeed, the liquid film remained stable until it was inflated by the pressurized shielding gas into a large bubble (see Fig. 7(a) at t = 8.3 ms) which eventually disintegrated (t = 8.9 ms), detaching the central part accompanied by numerous splashes.
Fig. 7. (a) Sequence of IR intensity images of an unsuccessful LDGf; (b) Solidified bubble form with a millimeter scale.
The above described outcome of the LDGw process using the Ag15Cu80P5 alloy foil was attributed to the properties of the Ag15Cu80P5 alloy. In particular, the Ag15Cu80P5 alloy thermal diffusivity D = O/(ȡ·cp) at 300 K was estimated to be D = 1.3·10-4 m2/s by calculating the heat conductivity O and specific heat cp by the rule of mixtures from the data for copper and silver. This thermal diffusivity is more than fifty times larger than the thermal diffusivity of Ni, which has a value of 2.5·10-5 m2/s. The large thermal diffusivity along with the relatively large thickness h = 0.2 mm could explain the rapid heat transfer in the Ag15Cu80P5 alloy foil from the irradiated ring into the center region. Another important property of the Ag15Cu80P5 alloy is the difference Tl – Ts between the liquidus and solidus temperatures, which is Tl – Ts = 156 K for the Ag15Cu80P5 alloy. At the intermediate temperatures, the
727
728
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
alloy is a mixture of a solid and liquid material which could explain the ability of the alloy to form large bubbles. This ability was strikingly demonstrated in one observed occasion, when at low laser power P and low overpressure p a bubble was formed and solidified into a cup-like structure as shown in Fig. 7(b). The irradiated spot diameter ds in this case was 2.5 mm, the square laser pulse power P was 3 kW, duration W = 20.0 ms, and the shielding gas overpressure p was 1.0 kPa. 4. Conclusions In this paper, the experimental setup and process of novel laser droplet generation from a metal foil (LDGf) are described along with the results of the initial experimental study of the LDGf process which included nickel and Ag15Cu80P5 alloy foils of different thicknesses. In the study, the process was characterized by a high speed IR camera. The experimental results showed that for a successful outcome of the process, i.e. detached droplet with repeatable properties and with minimal amount of splashes, the proper shape of the laser beam is annular, while the process using the Gaussian laser beam was not successful. The successful process was observed to consist of the following stages: melting of the ring-shaped region of the foil; downward acceleration of the solid central part; detachment of the central part by break-up of the molten ring around the solid central part; melting of detached central part while it travels through the laser beam focus; and droplet formation under the influence of the surface tension. The important process parameters identified were the laser beam shape, laser beam power profile P(t), shielding gas overpressure p, foil thickness h, and physical properties of the foil material. Using the experimental data, the influences of the process parameters were determined as follows. By selecting the outer diameter of the annular laser beam on the foil surface, the diameter of the generated droplet was directly defined via the volume of the melted material. The symmetry of the annular laser beam was found to be very important at thinner foils (Ni, thickness 0.0125 mm), while at thicker foils (Ni, thickness 0.1 mm) the process was not very sensitive to small asymmetry of the laser beam. The experimental results showed that such laser pulse power profile P(t) should be selected which would melt the ring shaped irradiated region of the foil as quick as possible while at the same time avoiding evaporation. Further, the laser pulse duration should be long enough to melt the material, detached from the foil, and form a droplet. The shielding gas overpressure p was found to mainly influence the detachment. At too high values of the overpressure p, the detachment was accompanied by splashes, while at too low values, the detachment was unstable or unsuccessful. A force balance theoretical model of threshold overpressure ptr was proposed, which results coincide well with the experimental data. Further, the experiments indicated the important influence of the material properties, especially the thermal diffusivity D of the foil material and the foil thickness h. When using a material with high thermal diffusivity, the foil thickness should be small and the ring shaped irradiated region of the foil should have large enough diameter to avoid premature melting of the centre region of the foil. In view of possible development of droplet deposition based technology, the usage of a Ni foil of thickness h = 0.1 mm to generate and deposit Ni droplets on titanium substrate was successfully demonstrated. In the future, a numerical model of the LDGf process will be constructed which would be able to scan a wider region of process parameters and to determine the dependencies of the generated droplet properties on the process parameters. Further experimental work will include consideration of different techniques for applying a time dependent laser pulse power and shielding gas overpressure. Acknowledgements This work was supported by ARRS – the Slovenian Research Agency research program P2-0241 and by the COST Action MP1106.
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
References Albert, F., Pfeiffer, C., Schmidt, M., Geiger, M., Flössel, M., Michaelis, A., 2011. Laser soldering and laser droplet joining for mechanical and electrical contacting of LTCC/PZT laminates. Journal of Laser Micro/Nanoengineering 6, 75-80. Chun, J., Passow, C. H., Suh, N. P., 1993. Droplet-based manufacturing. CIRP Annals - Manufacturing Technology 42, 235-238. Conway, P., Fu, E., Williams, K., Williams, D., 2002. Precision high temperature lead-free solder interconnections by means of high-energy droplet deposition techniques. CIRP Annals - Manufacturing Technology 51, 177-180. Dreizin, E. L., 1997. Droplet welding: A new technique for welding electrical contacts. Welding Journal 1997-04, 67-73. Govekar, E., Jeriþ, A., Weigl, M., Schmidt, M., 2009. Laser droplet generation: Application to droplet joining. CIRP Annals - Manufacturing Technology 58, 205-208. Held, C., Quentin, U., Heberle, J., Gürtler, F., Weigl, M., Schmidt, M., 2012. Laser droplet brazing for the electrical contacting of composite materials with integrated active elements. Physics Procedia 39, 585-593. Hoving, W., Jahrsdörfer, B., 2001. Laser droplet weld - Ein innovatives Fuegeverfahren Laser in der Elektronikproduktion & Feinwerktechnik, in “Laser in der Elektronikproduktion & Feinwerktechnik”. In: Geiger, M., Otto, A. (Eds.). Meisenbach Verlag, Bamberg, pp. 1-12. Jahrsdörfer, B., Esser, G., Geiger, M., Govekar, E., 2003. Laser droplet weld - an innovative joining technology opens new application possibilities. Photon Processing in Microelectronics and Photonics II, Proceedings of SPIE 4977, 518-529. Jeriþ, A., Grabec, I., Govekar, E., 2009. Laser droplet welding of zinc coated steel sheets. Science and Technology of Welding and Joining 14, 662-668. Jeriþ, A., 2010. Droplet formation by coaxial input of laser light into a metal wire. PhD Thesis, University of Ljubljana, Ljubljana, 2010. Jeromen, A., Held, C., Govekar, E., Roth, S., Schmidt, M., 2014. Modelling of droplet detachment in the laser droplet brazing process. Journal of Materials Processing Technology 214, 737-749. Kokalj, T., Klemenþiþ, J., Mužiþ, P., Grabec, I., Govekar, E., 2006. Analysis of the laser droplet formation process. Journal of Manufacturing Science and Engineering 128, 307-314. Kuznetsov A., Jeromen, A., Govekar E., 2014. Droplet detachment regimes in annular laser beam droplet generation from a metal wire. CIRP Annals - Manufacturing Technology, http://dx.doi.org/10.1016/j.cirp.2014.03.051, in press. Liu, H., 2000. Science and engineering of droplets: fundamentals and applications. Noyes Publications / William Andrew Publishing, Norwich, 2000. Qi, L., Chao, Y., Luo, J., Zhou, J., Hou, X., Li, H., 2012. A novel selection method of scanning step for fabricating metal components based on micro-droplet deposition manufacture. International Journal of Machine Tools & Manufacture 56, 50-58. Zarzalejo, L.J., Schmaltz, K.S., Amon, C.H., 1999. Molten droplet solidification and substrate remelting in microcasting, Part I: Numerical modeling and experimental verification. Heat and Mass Transfer 34, 477-485.
729
ScienceDirect Physics Procedia 56 (2014) 720 – 729
8th International Conference on Photonic Technologies LANE 2014
Laser droplet generation from a metal foil Andrej Jeromena,*, Alexander Kuznetsova, Edvard Govekara a
Faculty of Mechanical Engineering, University of Ljubljana, Aškerþeva 6, SI-1000 Ljubljana, Slovenia
Abstract Metal droplets are used in various droplet based innovative technologies, especially in joining of temperature sensitive components where the amount of applied energy and its spatial distribution are important. Laser light has already been proved to be a suitable energy source for metal droplet generation, since it enables high spatial and temporal control of the energy input. In this contribution, a novel process of laser droplet generation (LDG) from a metal foil is presented, which could provide increased flexibility of the generated droplet parameters compared to the existing LDG processes. In the novel LDG process, a horizontally placed foil is irradiated by a laser beam from above and the melted part of the foil is detached by shielding gas overpressure above the foil. The results of presented preliminary experimental study show that droplet formation from metal foil is possible by means of an annular laser beam. In addition to this, the space of process parameters is explored, showing influences of temporal intensity distributions of the annular laser beam, shielding gas overpressure, and foil thickness and material on the process outcome. © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and blind-review under responsibility of the Bayerisches Laserzentrum GmbH. Peer-review under responsibility of the Bayerisches Laserzentrum GmbH Keywords: laser droplet generation; metal foil; annular laser beam
1. Introduction Nowadays, metal droplets are used as a basis of various innovative technologies, e.g. Chun and Passow (1993) and Liu (2000), which are employed in different applications. Some of the reported applications, based on metal droplets, include joining of electrical contacts by Dreizin (1997) and Conway et al. (2002), joining of heat sensitive and dissimilar materials by Jahrsdörfer et al. (2003) and Govekar et al. (2009), welding of zinc coated steel sheets by
* Corresponding author. Tel.: +386-1-4771-514; fax: +386-1-251-8567 . E-mail address: [email protected]
1875-3892 © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Bayerisches Laserzentrum GmbH doi:10.1016/j.phpro.2014.08.079
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
721
Jeriþ et al. (2009), microcasting by Zarzalejo et al. (1999), 3-D structuring by Qi et al. (2012), and droplet brazing by Held et al. (2012) and Jeromen et al. (2014). For different applications, different droplet generation systems and processes have been developed. They could be divided into two groups: drop-on-demand, where individual droplets are separately generated, and continuous droplet generation, where a sequence of identical droplets is generated at a constant frequency. Drop-on-demand process is required in applications, where metal droplets need to be deposited individually to realize point-like joints, for example in electrical contacting of temperature-sensitive electrical and electronic components such as piezoelectric laminates, described by Albert et al. (2011). Originally, metal drop-on-demand generation process with solid metal as input material was introduced by Dreizin (1997) who generated droplets by melting a metal wire using an electric arc. Later, laser droplet generation from a wire (LDGw) has been described by Hoving and Jahrsdörfer (2001) and analyzed by Kokalj et al. (2006). In LDGw, the droplet is generated by melting the tip of a metal wire by a laser pulse. As laser light enables high spatial and temporal control of the energy input it proved to be a very suitable energy source for metal droplet generation. However, to reliably generate metal droplets from a wire, a symmetrical energy input by means of an annular laser beam, as well as a precise control and synchronization of laser pulse power and wire feeding are needed, as demonstrated by Kuznetsov et al. (2014). Recently, Held et al. (2012) reported on laser droplet brazing, where a spherical braze preform, blocking a tapered nozzle, is melted by the laser pulse and detached from the nozzle by the overpressure of shielding gas inside the nozzle. The technique is promising for electrical contacting applications, if the nozzle instability issue due to the thermal shocks is addressed adequately. As an alternative laser droplet generation process, a novel process of laser droplet generation from a foil (LDGf) is presented in this paper. The LDGf process has a potential to provide increased flexibility of the generated droplet parameters, in particular generation of smaller droplets, compared to the above laser droplet generation processes by using a less complex system. What is more, in LDGf, in contrast to laser droplet brazing system, no permanent system components are thermally loaded which implies higher stability of the system. In the following, the LDGf experimental system and process are described. Next, the results of the initial experimental study of LDGf, including two different laser beam shapes, two different foil materials and different foil thicknesses, are presented and discussed. In the conclusions, the main results of the study are gathered. 2. Experimental system and process of laser droplet generation from a foil The LDGf experimental system is presented in Fig. 1(a). It consisted of a Nd:Yag pulsed laser source with wavelength of 1064 nm and a Gaussian beam. Optionally, an optical system, designed by Jeriþ (2010), was used to shape the Gaussian laser beam into an annular laser beam (shown in Fig. 1(a)). The approximate caustic of the annular laser beam, determined by measuring the inner and outer diameter of the heat affected zone on irradiated metal surface at different positions along the beam axis, is presented in Fig. 1(b). Circles in Fig. 1(b) denote the outer and squares the inner diameter of the annular laser beam. The laser beam was guided to the surface of the metal foil. For this purpose, an experimental cell, pressurized by shielding gas, was designed as shown in Fig. 1(a). The top and the bottom opening of the cell were closed by a glass window, transparent for the laser light, and by the clamped metal foil. The overpressure p of argon shielding gas inside the cell was set manually and was monitored by a pressure transducer. The volume of the cell was approximately 4 cm3. To prevent oxidation, the argon atmosphere was maintained also below the foil. For the experimental process characterization, a high speed IR camera was used. The LDGf process consists phenomenologically of two consecutive phases: (1) laser beam selective melting of the foil, and (2) detachment of the melted material by the shielding gas overpressure with formation of the droplet due to surface tension. At given metal foil thickness h and its physical properties, the main process parameters are the spatial profile (shape) of the laser beam, irradiated spot diameter ds, laser pulse power time profile P(t) and shielding gas overpressure p. The foil material used in experiments was nickel of purity 99.0% and thickness h = 0.1 and 0.0125 mm. In view of potential brazing applications in electrical contacting of copper wires on silver contact pads, the Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was also selected as it is used for flux-free brazing of copper. The physical properties of both materials are gathered in Table 1. Instead of the unavailable data for Ag15Cu80P5, the values for copper and
722
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
silver are given. Compared to Ag15Cu80P5, Ni has a higher melting point and, considering values for copper and silver, much lower heat conduction, higher specific heat and higher viscosity.
Fig. 1. (a) Scheme of LDGf experimental system; (b) Annular laser beam caustic.
Table 1. Physical properties of used foil materials (from Liu (2000) except where indicated). Property
Ni
Ag15Cu80P5
Density ȡ at 300 K [kg/m3]
7905
8442 (product data sheet)
Temperature of melting Tm [K]
1728
917 (Ts), 1073 (Tl) (product data sheet)
Thermal conductivity O at 300 K [W/(m·K)]
88.5
397 (Cu), 425 (Ag)
Specific heat cp at 300 K [J/(kg·K)]
452
386 (Cu), 234 (Ag)
Viscosity ȝ at melting point [Pa·s]
1.7·10-4
3.0·10-4 (Cu), 4.5·10-4 (Ag)
Surface tension V at melting point [N/m]
1.8
1.3 (Cu), 0.90 (Ag)
3. Experiments, results and discussion In the following, the results of the initial LDGf experiments and the influence of the main process parameters on the process outcome are presented and discussed. In experiments, the Gaussian and the annular laser beam were used at different diameters ds of irradiated spot. In the experiments also different values of the shielding gas overpressure p were applied, and nickel and Ag15Cu80P5 alloy foils of selected thicknesses were used. The process was characterized by the high speed IR camera with frame of 96x80 pixels, integration time of 6 ȝs and frame rate of 1660 fps. 3.1. Gaussian laser beam First, the LDGf experiments were performed using the Gaussian laser beam, which was selected as the most common laser beam profile. To select different values of irradiated spot diameter ds, the experimental cell with the foil was positioned at the corresponding position above the laser beam focus. By the foil being positioned above the
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
723
laser beam focus, additional heating of the detached foil material, which travelled downwards through the laser beam focus, was realized. In the experiments, Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was used with two irradiated spot diameter ds values ds = 1.0 and 2.0 mm. At ds = 1.0 mm, a square laser pulse of power P = 3.0 kW and duration W = 6.5 ms and at ds = 2.0 mm a square pulse of P = 3.9 kW and W = 20.0 ms were used. The threshold shielding gas overpressure ptr for detachment of the melted part of the foil at ds = 1.0 and 2.0 mm was experimentally determined to be ptr = 8.5 kPa and 5.0 kPa, respectively. It was observed from the recorded sequences of the IR intensity images that at overpressures p above the threshold overpressure ptr no successful LDGf, resulting in a droplet, was realized. In all observed cases, the melted region either broke up and the melted material was pulled to the edge or it inflated into a bubble which subsequently disintegrated into splashes, allowing the pressurized shielding gas to escape from the cell. An example sequence of recorded IR intensity images at ds = 2.0 mm, showing unsuccessful LDGf with bubble inflation and disintegration, is presented in Fig. 2. Indicated time t below images is related to the time of the laser pulse trigger signal. In the images, brighter color denotes higher temperatures. In the leftmost image of Fig. 2, the irradiated spot is visible. After the irradiated part of the foil material was melted, the melted region was inflated into a bubble at t = 19.3 – 19.9 ms which then detached at t | 20.5 ms. After detachment, the bubble disintegrated into splashes although the laser beam was turned off at t = 20.0 ms. At the top of the rightmost two images, the circular opening that remained in the foil after detachment of the irradiated foil spot is clearly visible.
Fig. 2. Sequence of IR intensity images of an unsuccessful LDGf.
In the performed experiments with the Gaussian laser beam no successful droplet generation was observed. The process resulted into break-up of the melted region with the melt being pulled to the edges of the opening or into the bubble formation and disintegration as well as. Both outcomes were attributed to the Gaussian laser beam intensity distribution, where the highest intensity is in the center of the irradiated spot. This caused the foil in the centre of the irradiated spot to melt first and thus gave rise to the above described behavior which results in unsuccessful process outcome. 3.2. Annular laser beam To avoid melting of the irradiated region in its centre, the annular laser beam, with a ring like shape of the irradiated foil surface region, was employed. As in the Gaussian laser beam case, the experimental cell was positioned above the laser beam focus, to select the value of the irradiated spot diameter ds. This layout also enabled heating of the detached material below foil by the laser beam. The annular laser beam caustic (see Fig. 1(b)) together with the geometry of the experimental cell allowed the available values of beam outer diameter dout, which is equal to the irradiated spot diameter ds, between 1.5 and 3.0 mm. Below dout = 1.5 mm, the beam shape was not annular but full (see Fig. 1(b)), and above dout = 3.0 mm, the beam was too wide to fit the experimental cell opening. In experiments, both Ni and Ag15Cu80P5 foils were used, the former with thicknesses h = 0.1 mm and 0.0125 mm and the latter with thickness h = 0.2 mm. 3.2.1. Nickel foil with thickness of 0.1 mm For LDGf with Ni foil with thickness h = 0.1 mm, three values of irradiated spot diameter ds were selected: ds = 1.5, 2.0, and 3.0 mm. Although the LDGf using the square laser pulse profile was already successful, the results shown in the following were obtained using the laser pulse profile P(t) with decreasing power. With such profile,
724
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
faster melting of the foil without evaporation was achieved. At ds = 1.5 mm, the pulse power P linearly decreased from 8 kW to zero in duration of the pulse W = 3 ms. At ds = 2.0 mm, the pulse power P linearly decreased from 8 kW to zero in W = 5 ms. At ds = 3.0 mm, the area of the irradiated surface was so large and the corresponding laser light intensity so low, that in order to provide the necessary energy for melting, the laser pulse consisted of a constant part with P = 7.5 kW and W = 5 ms followed by a linear decrease from P = 7.5 kW to zero in the following millisecond. At the above described parameters, the second phase of LDGf process – detachment of the melted part and formation of the droplet – was observed to consist of three consecutive steps. The observed steps are presented by the sequence of IR intensity images in Fig. 3. Indicated time t below the images is related to the time of the laser pulse trigger signal. The second phase of LDGf process started after the ring shaped irradiated region of the foil was melted at t | 3.5 ms (see Fig. 3). After this, at time t = 4.7 ms, the central circular part, which was still solid, was pressed downwards due to the deformation of the molten ring by the shielding gas overpressure p. Then, the solid central part was detached by the break-up of the molten ring (t | 5.3 ms) which had connected the central part with the foil. Finally, the still solid central part, travelling downwards through the laser beam focus, was melted by the laser light and under influence of the surface tension, a droplet was formed (t = 10.7 ms).
Fig. 3. Sequence of IR intensity images of a successful LDGf at ds = 3.0 mm and p = 1.7 kPa.
The above described detachment and formation of the droplet was observed to be very repeatable and accompanied with minor or no splashes. However, depending on the shielding gas overpressure p and the irradiated spot diameter ds, different regimes of the LDGf process, leading to different outcomes, were observed experimentally. The regimes are presented in a form of an overpressure p versus diameter ds diagram, shown in Fig. 4. In the diagram, stars denote the values of p and ds where the droplet generation was observed to be optimal, i.e. repeatable and with minimum amount of splashes. At increased overpressure p, the amount of splashes, produced by the ring break-up, increased, which is denoted by squares in the diagram. At slightly lower than optimal overpressure p values, denoted by circles, the detachment was observed to become unstable. The inspection of the acquired IR image sequences showed that the instability was caused by the incomplete and un-simultaneous breakup of the molten ring which resulted in the detached material falling at an angle and, therefore, missing the laser beam focus. The described instability was attributed to the imperfect spatial symmetry of the annular laser beam which caused uneven heating and melting of the irradiated ring-shaped region of the foil. At further decreased values of the shielding gas overpressure p, denoted by crosses, the detachment did not occur at all. In those cases the melted region of the foil either remained partly or completely attached to the solid edge of the foil and resolidified.
725
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
Fig. 4. Shielding gas overpressure p versus irradiated spot diameter ds. The symbols denote different detachment regimes. Theoretical gas overpressure threshold ptr dependence is denoted by solid line.
The threshold shielding gas overpressure ptr for reliable detachment of the melted part of the foil at ds = 1.5, 2.0, and 3.0 mm was experimentally determined to be ptr = 5.5, 3.5, and 1.7 kPa, respectively. These values are denoted by star symbols in Fig. 4. By considering a simple force balance model, the threshold overpressure ptr dependence on the irradiated spot diameter ds was determined also theoretically. The model describes the force balance of the central circular part at the moment directly before detachment, when the central circular part is connected with the foil by the liquid in a shape of a circular wall (see Fig. 3 at t = 4.7 ms). In the model, the force of gravity Fg and the overpressure force Fp, acting downwards on the central circular part, are balanced with the surface tension force FV which pulls the central part upwards. The force of gravity was expressed by Fg = Sds2hȡg, where h is the foil thickness, ȡ the density of the foil material, and g the gravitational acceleration. The overpressure force was expressed by Fp = Sds2ptr/4, and the surface tension force by FV = 2SdsV, where V is the surface tension of liquid material and where both (inner and outer) surfaces of the circular wall were taken into account. The threshold overpressure ptr dependence on the diameter ds is then determined from the force balance equation Fg + Fp = FV: ptr (d s )
8V hU g . ds
(1)
It can be shown that the second term on the right, resulting from the force of gravity, is negligible since it is three orders of magnitude smaller than the first term at parameter values used in the experiments. This shows that the gravity has a negligible effect on the LDGf process. The ptr(ds) dependence, expressed by Eq. (1), is presented in Fig. 4 as a solid line and corresponds well to the experimentally determined threshold shielding gas overpressure ptr values, denoted by stars. The generated droplet diameter in general depends on the diameter ds of irradiated spot. At given diameter ds, the diameter dex of the experimentally generated droplets was determined by measuring generated droplets which solidified in a free fall. The measurement error due to the imperfect spherical shape of the solidified droplets was estimated to be 0.1 mm. The measured dex values at the selected values of the irradiated spot diameter ds are gathered in Table 2. There they are compared to the theoretically determined droplet diameter dth values, calculated from the volume of the melted spot with diameter ds of the foil by equation: d th
3
3d s 2 h . 16
(2)
It can be seen from Table 2 that the experimental values are equal to the theoretical values which indicates that, within the measurement error, the complete volume of the melted material was detached in the form of a droplet.
726
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729 Table 2. Measured dex and calculated dth generated droplet diameters at selected irradiated spot diameter ds values. Irradiated spot diameter ds [mm]
measured dex [mm]
theoretical dth [mm]
1.5
0.7 ± 0.1
0.70
2.0
0.85 ± 0.1
0.84
3.0
1.1 ± 0.1
1.11
In a view of potential joining applications, deposition of generated droplets was demonstrated on a titanium substrate of 0.6 mm thickness which was placed 6 mm below the bottom surface of the foil. In Fig. 5, photographs of the top and the side view of the deposited droplets at the selected irradiated spot diameter ds values are shown. Despite the relatively large distance between the foil and the substrate, the generated droplets contained enough energy to wet well the substrate surface.
Fig. 5. Top and side view photographs of deposited droplets at the selected values of the irradiated spot diameter ds.
3.2.2. Nickel foil with thickness of 0.0125 mm To be able to produce droplets of smaller diameter, the LDGf process was performed using thinner Ni foil with thickness h = 0.0125 mm at the irradiated spot diameter ds values of 1.5, 2.0, and 3.0 mm. At this foil thickness the experiments of the LDGf were unsuccessful at all values of the shielding gas overpressure p, regardless of the irradiated spot diameter ds and the laser pulse power profile P(t). At lower values of the shielding gas overpressure p, including p = 0, the melted part of the foil was immediately pulled by the surface tension to the edge of the opening. At increased overpressure p values, the LDGf process resulted in unstable detachment. With further increase of the overpressure p, the droplet detachment phase was accompanied with splashes and with too high acceleration of the central part to be melted while travelling through the laser beam focus. Example IR intensity sequences of the corresponding LDGf processes with no detachment and with unstable detachment are shown in Fig. 6. Indicated time t below images is related to the time of the laser pulse trigger signal. In both cases, the irradiated spot diameter ds was 2.0 mm, and the laser pulse power P linearly decreased from 2 kW to zero in W = 1.2 ms. In the first case, shown in Fig. 6(a), the overpressure p was set to 2.0 kPa. The IR images show the melted material being pulled to the edge, accompanied by some splashes. In attempt to achieve a stable detachment, the overpressure p was increased to 6.0 kPa, which is nearly twice the optimal value as determined by the theoretical model. As shown by the acquired IR images in Fig. 6(b), the detaching central part was held at one side (see image at t = 2.3 ms), detached at a large angle with respect to the vertical axis, and completely missed the laser beam focus. The main reason for the observed behavior was presumably the annular laser beam asymmetry. The beam asymmetry is evident in the first image (at t = 1.1 ms) in both Fig. 6(a) and (b), where beside the ring shaped heated region of the foil also a part of the inner circular region was heated. Due to the small thickness of the foil, the shape of the heated region approximately corresponds to the shape of the laser beam. The observed asymmetry of the laser beam was a consequence of imperfect alignment of the beam shaping optical components. Due to the resulting asymmetric melting, the central part was pulled to a side by the surface tension. The small thickness of the foil contributed considerably to the described behavior, as the melted region was evidently unstable and broke immediately. Its width to thickness ratio was approximately 50 in the case of foil thickness h = 0.0125 mm, compared to approximately 5 in the case of h = 0.1 mm, where the ring shaped liquid region was more stable and
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
broke-up only when it was thinned enough by stretching due to the central part being pressed downwards by the overpressure p.
Fig. 6. Sequence of IR intensity images of an unstable detachment at (a) p = 2.0 kPa and (b) p = 6.0 kPa.
3.2.3. Ag15Cu80P5 alloy with thickness of 0.2 mm In view of potential brazing applications the Ag15Cu80P5 alloy foil of thickness h = 0.2 mm was employed in LDGf. As above, the irradiated spot diameter ds values of 1.5, 2.0, and 3.0 mm were selected while the corresponding laser power profiles P(t) and shielding gas overpressure p values were attempted to be experimentally determined in order to achieve a successful LDGf process. The acquired IR image sequences showed that at irradiated spot diameter ds values of 1.5 and 2.0 mm, the process outcome was very similar to the case with Gaussian beam. Despite the ring shaped irradiated region, the central region melted virtually at the same time as the irradiated region, which caused inflation of the melted region into a bubble and its disintegration into splashes, as described in Section 3.1. In attempt to avoid premature melting of the central region before the beginning of the detachment phase, the pulse power was increased to decrease time needed for melting the ring shaped irradiated region. However, before reaching fast enough melting of the irradiated region by the increased laser pulse power, the material at the irradiated surface started to evaporate, causing violent splashes. Only in the case of the irradiated spot diameter ds = 3.0 mm it was possible to melt the ring shaped region of the foil before the central region melted. Such a case, with the laser pulse power P linearly decreasing from 8 kW to zero in W = 8.0 ms and the shielding gas overpressure p = 2.0 kPa, is presented in Fig. 7(a). Indicated time t below images is related to the time of the laser pulse trigger signal. The images in Fig. 7(a) show the downward acceleration of the central part. However, the liquid, connecting the central part to the foil, did not break-up as fast as in the corresponding case with Ni (see Fig. 3). Indeed, the liquid film remained stable until it was inflated by the pressurized shielding gas into a large bubble (see Fig. 7(a) at t = 8.3 ms) which eventually disintegrated (t = 8.9 ms), detaching the central part accompanied by numerous splashes.
Fig. 7. (a) Sequence of IR intensity images of an unsuccessful LDGf; (b) Solidified bubble form with a millimeter scale.
The above described outcome of the LDGw process using the Ag15Cu80P5 alloy foil was attributed to the properties of the Ag15Cu80P5 alloy. In particular, the Ag15Cu80P5 alloy thermal diffusivity D = O/(ȡ·cp) at 300 K was estimated to be D = 1.3·10-4 m2/s by calculating the heat conductivity O and specific heat cp by the rule of mixtures from the data for copper and silver. This thermal diffusivity is more than fifty times larger than the thermal diffusivity of Ni, which has a value of 2.5·10-5 m2/s. The large thermal diffusivity along with the relatively large thickness h = 0.2 mm could explain the rapid heat transfer in the Ag15Cu80P5 alloy foil from the irradiated ring into the center region. Another important property of the Ag15Cu80P5 alloy is the difference Tl – Ts between the liquidus and solidus temperatures, which is Tl – Ts = 156 K for the Ag15Cu80P5 alloy. At the intermediate temperatures, the
727
728
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
alloy is a mixture of a solid and liquid material which could explain the ability of the alloy to form large bubbles. This ability was strikingly demonstrated in one observed occasion, when at low laser power P and low overpressure p a bubble was formed and solidified into a cup-like structure as shown in Fig. 7(b). The irradiated spot diameter ds in this case was 2.5 mm, the square laser pulse power P was 3 kW, duration W = 20.0 ms, and the shielding gas overpressure p was 1.0 kPa. 4. Conclusions In this paper, the experimental setup and process of novel laser droplet generation from a metal foil (LDGf) are described along with the results of the initial experimental study of the LDGf process which included nickel and Ag15Cu80P5 alloy foils of different thicknesses. In the study, the process was characterized by a high speed IR camera. The experimental results showed that for a successful outcome of the process, i.e. detached droplet with repeatable properties and with minimal amount of splashes, the proper shape of the laser beam is annular, while the process using the Gaussian laser beam was not successful. The successful process was observed to consist of the following stages: melting of the ring-shaped region of the foil; downward acceleration of the solid central part; detachment of the central part by break-up of the molten ring around the solid central part; melting of detached central part while it travels through the laser beam focus; and droplet formation under the influence of the surface tension. The important process parameters identified were the laser beam shape, laser beam power profile P(t), shielding gas overpressure p, foil thickness h, and physical properties of the foil material. Using the experimental data, the influences of the process parameters were determined as follows. By selecting the outer diameter of the annular laser beam on the foil surface, the diameter of the generated droplet was directly defined via the volume of the melted material. The symmetry of the annular laser beam was found to be very important at thinner foils (Ni, thickness 0.0125 mm), while at thicker foils (Ni, thickness 0.1 mm) the process was not very sensitive to small asymmetry of the laser beam. The experimental results showed that such laser pulse power profile P(t) should be selected which would melt the ring shaped irradiated region of the foil as quick as possible while at the same time avoiding evaporation. Further, the laser pulse duration should be long enough to melt the material, detached from the foil, and form a droplet. The shielding gas overpressure p was found to mainly influence the detachment. At too high values of the overpressure p, the detachment was accompanied by splashes, while at too low values, the detachment was unstable or unsuccessful. A force balance theoretical model of threshold overpressure ptr was proposed, which results coincide well with the experimental data. Further, the experiments indicated the important influence of the material properties, especially the thermal diffusivity D of the foil material and the foil thickness h. When using a material with high thermal diffusivity, the foil thickness should be small and the ring shaped irradiated region of the foil should have large enough diameter to avoid premature melting of the centre region of the foil. In view of possible development of droplet deposition based technology, the usage of a Ni foil of thickness h = 0.1 mm to generate and deposit Ni droplets on titanium substrate was successfully demonstrated. In the future, a numerical model of the LDGf process will be constructed which would be able to scan a wider region of process parameters and to determine the dependencies of the generated droplet properties on the process parameters. Further experimental work will include consideration of different techniques for applying a time dependent laser pulse power and shielding gas overpressure. Acknowledgements This work was supported by ARRS – the Slovenian Research Agency research program P2-0241 and by the COST Action MP1106.
Andrej Jeromen et al. / Physics Procedia 56 (2014) 720 – 729
References Albert, F., Pfeiffer, C., Schmidt, M., Geiger, M., Flössel, M., Michaelis, A., 2011. Laser soldering and laser droplet joining for mechanical and electrical contacting of LTCC/PZT laminates. Journal of Laser Micro/Nanoengineering 6, 75-80. Chun, J., Passow, C. H., Suh, N. P., 1993. Droplet-based manufacturing. CIRP Annals - Manufacturing Technology 42, 235-238. Conway, P., Fu, E., Williams, K., Williams, D., 2002. Precision high temperature lead-free solder interconnections by means of high-energy droplet deposition techniques. CIRP Annals - Manufacturing Technology 51, 177-180. Dreizin, E. L., 1997. Droplet welding: A new technique for welding electrical contacts. Welding Journal 1997-04, 67-73. Govekar, E., Jeriþ, A., Weigl, M., Schmidt, M., 2009. Laser droplet generation: Application to droplet joining. CIRP Annals - Manufacturing Technology 58, 205-208. Held, C., Quentin, U., Heberle, J., Gürtler, F., Weigl, M., Schmidt, M., 2012. Laser droplet brazing for the electrical contacting of composite materials with integrated active elements. Physics Procedia 39, 585-593. Hoving, W., Jahrsdörfer, B., 2001. Laser droplet weld - Ein innovatives Fuegeverfahren Laser in der Elektronikproduktion & Feinwerktechnik, in “Laser in der Elektronikproduktion & Feinwerktechnik”. In: Geiger, M., Otto, A. (Eds.). Meisenbach Verlag, Bamberg, pp. 1-12. Jahrsdörfer, B., Esser, G., Geiger, M., Govekar, E., 2003. Laser droplet weld - an innovative joining technology opens new application possibilities. Photon Processing in Microelectronics and Photonics II, Proceedings of SPIE 4977, 518-529. Jeriþ, A., Grabec, I., Govekar, E., 2009. Laser droplet welding of zinc coated steel sheets. Science and Technology of Welding and Joining 14, 662-668. Jeriþ, A., 2010. Droplet formation by coaxial input of laser light into a metal wire. PhD Thesis, University of Ljubljana, Ljubljana, 2010. Jeromen, A., Held, C., Govekar, E., Roth, S., Schmidt, M., 2014. Modelling of droplet detachment in the laser droplet brazing process. Journal of Materials Processing Technology 214, 737-749. Kokalj, T., Klemenþiþ, J., Mužiþ, P., Grabec, I., Govekar, E., 2006. Analysis of the laser droplet formation process. Journal of Manufacturing Science and Engineering 128, 307-314. Kuznetsov A., Jeromen, A., Govekar E., 2014. Droplet detachment regimes in annular laser beam droplet generation from a metal wire. CIRP Annals - Manufacturing Technology, http://dx.doi.org/10.1016/j.cirp.2014.03.051, in press. Liu, H., 2000. Science and engineering of droplets: fundamentals and applications. Noyes Publications / William Andrew Publishing, Norwich, 2000. Qi, L., Chao, Y., Luo, J., Zhou, J., Hou, X., Li, H., 2012. A novel selection method of scanning step for fabricating metal components based on micro-droplet deposition manufacture. International Journal of Machine Tools & Manufacture 56, 50-58. Zarzalejo, L.J., Schmaltz, K.S., Amon, C.H., 1999. Molten droplet solidification and substrate remelting in microcasting, Part I: Numerical modeling and experimental verification. Heat and Mass Transfer 34, 477-485.
729